The following article was published in ASHRAE Journal, February 2009.

Copyright 2009 American Society of Heating, Refrigerating and AirConditioning Engineers, Inc. It is presented for educational purposes only. This article may not be copied and/or distributed electronically or in
paper form without permission of ASHRAE.

Frost

On Air-Cooling Evaporators
By Douglas T. Reindl, Ph.D, P.E., and Todd B. Jekel, Ph.D, P.E.

n this article, we focus on the operation of air-cooling evaporators in industrial

refrigeration applications at temperatures below freezing. These heat exchangers

are generally applied to control the environmental conditions in holding freezers,

dynamic blast freezers, stationary blast cells, refrigerated docks, as well as other
lower temperature conditioned spaces found in food manufacturing and distribution
facilities. We review the factors that influence the formation of frost on the evaporator surfaces and discuss how frost accumulation impacts coil performance.
Air-cooling evaporators are refrigerantto-air heat exchangers widely used in
industrial refrigeration, commercial refrigeration, and heat pump systems. Known
as forced-circulation air coolers1 or aircooling evaporators, these heat exchangers use tubes to carry refrigerant with fins
February 2009

applied to the tube exterior. Individual

tubes of the heat exchanger are arranged
in multiple rows of parallel circuits to
achieve increased thermal performance.
Refrigerant evaporates inside the tubes as
it absorbs heat from air flowing over the
outside surface of the finned tubes.

When air-cooling evaporators operate

with both coil surface temperatures below
32F (0C) and entering air dew-point
temperatures above the coil surface temperature, moisture from the air being cooled
will accumulate on the fins and tubes of
the coil in the form of frost. The formation
and growth of frost on the evaporator will
decrease the coils cooling capability.
What is the root cause of capacity decrease
from frosted coil?
a. Increased air-side pressure
drop; thereby, decreasing
airflow through the coil.
b. Increased resistance to heat
transfer between the air and the
About the Authors
Douglas T. Reindl, Ph.D., P.E., is a professor
and director and Todd B. Jekel, Ph.D., P.E., is
assistant director at the University of WisconsinMadisons Industrial Refrigeration Consortium in
Madison, Wis.

ASHRAE Journal

27

refrigerant due to the insulating effects of the frost.

c. Both of the above.
d. None of the above.
If you answered c, you are correct. A number of papers
highlighting this phenomena have been published.2,3,4 Aljuwahel, et al,5 confirmed that the single greatest factor reducing
evaporator capacity due to frost accumulation is the decrease
in airflow rate due to its effect on air-side pressure drop as
originally suggested by Stoecker.2

provides some basic guidance for typical fin spacing on coils

over a range of operating conditions.
Table 1 provides some basic guidance for typical fin spacing
on coils over a range of operating conditions.
It is important to work with your evaporator manufacturer
to select an appropriate coil once you understand the operating
environment for that coil.
Coil Location. Coils located in regions with supersaturated
moisture will experience a much higher rate of frost accumulation and plugging. Figure 2 shows an example of warm, moist
Factors Influencing Coil Capacity During Frosting
air from a dock infiltrating into a freezer. Because the infiltrating
What are the factors that control how fast my evaporators air is more buoyant, it rises to the ceiling. The rapid cooling of
capacity will decrease due to frosting?
the supersaturated moist infiltrating air causes crystals of ice to
A number of factors influence the rate of frost
form in the air rather than on the coil surface. If
accumulation on a coil resulting in increased
an evaporator is located immediately above the
air-side pressure drop and reduced airflow rate
door opening, it will see a significant load of
supersaturated moisture and will quickly plug the
through the coil.
face of the coil with frost. The accelerated rate of
Fin spacing. Sometimes referred to as
frost accumulation on the entering side of the coil
fin pitch, the density of fins applied to an
surface is due to impaction and interception of
evaporator will have a dramatic effect on the
the ice crystals onto the coil surface. If the moist
susceptibility of coil frosting and the resulting
dock air had an opportunity to blend with the
capacity loss due to blocking airflow. Although
colder drier freezer air, the rate of frost-induced
increased fin density (decreased fin spacing)
plugging on the coil would be reduced.
may be desirable because it increases the surMoisture Load. In applications with sigface area available for heat transfer, the reduced
nificant moisture loading, the rate of frost acspacing between fins will result in a decrease of
Figure 1: An evaporator coil with
cumulation can rise dramatically, accelerating
the open area available for air to flow as frost
the first two tube rows fin free.
the capacity loss. The severity of the moisture
accumulates on the coil. A coil design with
narrow fin spacing will plug with frost more rapidly, requir- load can be characterized in terms of the sensible heat ratio
ing more frequent defrosts. A balance between fin spacing that (SHR). The sensible heat ratio represents the ratio of the space
provides adequate coil heat transfer surface area, but that does sensible load to the space total load as given by:
not lead to a rapid buildup of frost, is required.
In low-temperature freezing systems, some coil designs have


(1)
fin spacing that varies from the air-entering side of the coil to
the air-leaving side of the coil to more effectively manage the
effects of frost accumulation. Figure 1 shows an evaporator
The term Qsensible represents that portion of a heat load that
for a low-temperature blast freezer where the first two rows causes the air temperature to rise while Qlatent is the portion
of the coil are bare tubes (no fins). Because moisture is being of the heat load attributable to a space moisture increase. As
extracted from the airstream as it moves from the entering to the moisture or latent load in a temperature-controlled space
the leaving side of the evaporator, successive rows in this coil increases, the sensible heat ratio decreases and operating
have increased fin density to provide surface area while miti- evaporators will experience increased difficulty removing the
gating the plugging effects of the frost accumulation. Table 1 moisture needed to meet the latent load. This is especially
Operating Temperature Range

Moisture/Frost Load

Typical Fin Pitch

(Fins per in.)

Comments

25F (32C) and Colder

HeavyModerate

03

Consider a Variable Fin Pitch Coil

Consider a Variable Fin Pitch Coil

Heavy

03

ModerateLight

23

HeavyModerate

Light

HeavyLight

46

25F (32C) to +10F (12C)

+10F (12C) to +35F (2C)
+35F (2C) to +50F (10C)

Avoid using the high fin density coils in applications

with airborne particulates (e.g., packaging areas).

Table 1: Review of typical evaporator fin spacing over a range of space operating conditions.
28

ASHRAE Journal

a s h r a e . o r g

February 2009

February 2009

ASHRAE Journal

Humidity Ratio

Dock Condition

relevant for those evaporators operating

in low-temperature environments (below
Frost
Accumulation
32F [0C]).
Ice Crystals
A number of past investigators have characterized envelopes of operating conditions
that lead to moisture and frost problems in
conditioned spaces. Figure 3 shows a series
of three process lines on a section of a lowLoading Dock
temperature psychrometric chart. The situation is typical of what happens when higher
temperature and humidity air from a dock
space infiltrates to a lower temperature and
Warm Inflow
humidity storage freezer space in an airflow
Freezer
pattern similar to that shown in Figure 2.
The psychrometric processes involved
with this situation are illustrated in three
Fog
Cold Outflow
separate situations shown in Figure 3. In
Iced Floor
each case the infiltrating air from the dock
is at a constant dry-bulb temperature of Figure 2: Airflow patterns from a dock to a freezer.14
50F (10C), while the relative humidity
0.012
varies from 60% (Case 1) to 30% (Case
3), which mixes with a freezer with a
Saturation Line
space temperature of 10F (12C).
0.010
The lowest surface temperature in the
freezer space is the evaporator coil, which
0.8 0.008
operates at 10F (23C)a point
commonly referred to as the apparatus
0.6 0.006
40F
dew-point temperature (ADP).
Case 1: Unfavorable Process Line
Mixed
Air
When the air from the dock enters into
Case 2: Borderline Favorable Process Line
30F
the freezer, a mixing process occurs. As
0.4 0.004
Case 3: Favorable Process Line
the warmer dock air blends with the cold20F
er air within the freezer, the mixed-air
0.2 0.002
ADP
10F
0F
condition progressively moves down the
10F
tie line that connects the dock condition
0.000
to the freezer condition. Case 1 illustrates 20
10
0
10
20
30
40
50
60
Dry-Bulb Temperature (F)
an unfavorable process line due to portions of the process line that connects Figure 3: Psychrometric chart showing process lines that range from unfavorable to favorthe mixed-air condition to the coil ADP able conditions for frosting.3
being above the saturation curve, which
leads to a supersaturated moisture condition where ice crystals a borderline favorable coil frosting condition. Case 3 is a lower
will form in the air. Smith6,7 identified the unfavorable frost dock humidity condition (30%) with a similar dock-freezer air
condition and noted how the presence of this condition adversely mixing ratio as Case 2. In this case, the process line stays below
affected evaporator performance, as well as causing significant the saturation curve yielding a favorable frost condition. We
icing effects on other cold surfaces within the freezer. Case 1 should note that the assumed freezer set temperature of 10F
would be reflective of entering air condition for an evaporator (12C) and the temperature difference between the freezer
located in close proximity to the dock-freezer doors where and coil refrigerant temperature (approximately the ADP) of
the coils entering air (a mixed-air condition) is closer to the 20F (11C) are both quite high. These conditions were used
dock condition due to the relatively narrow opportunity of the to illustrate the concepts of unfavorable and favorable frostinfiltrating air to fully mix with the freezer air.
ing conditions. Set temperatures for freezers range from 5F
In Case 2, the dock relative humidity is lower (approximately (20C) to 20F (29C), while a 10F (6C) temperature
40%), and the coil entering air is much closer to the freezer difference is more typical.
condition, indicating that a more thorough mixing process has
How can I minimize or avoid conditions that lead to unfavoroccurred as would be the case for evaporators located further able frosting?
away from the dock-freezer doors. In this case, the process line
Obviously, the source of the moisture should be identified and
is approaching, but not crossing, the saturation curve leading to minimized. In cases where moisture originates from infiltrating
29

Humidity Ratio

air, appropriate techniques to reduce that

0.012
infiltration rate should be pursued. This
Saturation Line
may require repairing seals, maintaining
0.010
door control, reducing openings (e.g., for
conveyors), and ensuring an appropriate
0.8
pressure balance between spaces at dif0.008
fering conditions. Further information on
these strategies can be found in several
0.6
0.006
40F
papers.3,8,9
Process Line
If the source of moisture is from prodOriginal
Dock
Condition
With Lower
0.4
Dock Set
ucts that are being processed, consider
30F
0.004
Unfavorable
Temperature
alternative means that can reduce moisProcess Line
20F
Revised Dock
ture loss. Strategies for reducing product
0.2
Temperature
0.002
moisture loss can include: packaging
ADP 10F
10F
0F
prior to cooling/freezing; pre-cooling
product; and crust-freeze product using
0.000
a cryogenic fluid prior to finish-freezing 20
10
0
10
20
30
40
50
60
Dry-Bulb Temperature (F)
with a mechanical freezing system (flash
freezing will create a crust on the product Figure 4: Psychrometric chart showing process lines for a lower dock set temperature to
surface to minimize desiccation). Tech- improve favorability of the freezer coil frosting condition.
niques to reduce moisture loss have the
added benefit of increasing product yield. Yield savings almost ing systems, the extent of unfavorable frost conditions can be
certainly will far outweigh energy cost benefits from reduced minimized but not eliminated.
moisture loads on evaporators. For many low-temperature freezFor spaces such as holding freezers, one common approach

Advertisement formerly in this space.

30

A S H R A E J o u r n a l

February 2009

for avoiding unfavorable frost conditions

is to lower the setpoint temperature of the
dock to increase the level of moisture removal at a higher evaporator temperature
(when compared to the freezer). Reducing the dock setpoint temperature to
something in the range of 35F (2C),
will permit air defrosting while providing significantly more moisture removal
when compared to a 50F (10C) space
setpoint. In some cases, hot-gas reheat is Figure 5: Low density frost forming on an evaporator due to high coil temperature difference
added at the dock evaporator to further and presence of supersaturated air.
increase the space sensible heat ratio.
Figure 4 shows a dock maintained at a 35F (2C) dry-bulb capacity decrease due to airflow blockage. Unfavorable frosttemperature with a relative humidity of 60%. A process line ing conditions leads to the formation of ice crystals directly
from the entering coil condition to the ADP shows it to be just in the airstream. There is a tendency for these ice crystals to
favorable as it approaches but does not cross the saturation precipitate onto cold surfaces within the space; however, they
curve. Adding reheat at the dock door can further increase the will ride along on air currents created by operating evaporator
dry-bulb temperature of infiltrating air, driving the process to fans. The frost crystals will readily adhere to the coil surface
an increasingly favorable frost condition process line.
by physical impaction or interception; thereby, blocking airflow.
Cleland3 offers other strategies for avoiding unfavorable Figure 5 shows the structure of unfavorable frost adhering to
frosting conditions but rightly places a particular emphasis on the surfaces of a variably finned, low-temperature evaporator
preventing the infiltration using door protection devices.
freezing unpacked product and operating with a moderately
Frost Type. Somewhat related to discussions in the previous high TD (difference in temperature between the entering air
section, the type of frost has an influence on the rate of coil and the evaporating refrigerant). In this case, the structure of

Advertisement formerly in this space.

February 2009

Advertisement formerly in this space.

ASHRAE Journal

31

the frost is extremely light and fluffy with minimal bonding to

the coil surface. We postulate that, in this case, the coil plugged
where the fins started and that the light, fluffy frost grew after
the coil blockage. As mentioned previously, this type of frost
degrades coil performance more rapidly than a higher density
frost as shown in Figure 6.
The higher density frost forming on the coil shown in Figure
6 occurs in spaces with favorable frosting conditions. Due to
the lack of supersaturated air, as well as an operating coil TD
of 10F (6C), moisture from the air forms on the coil by a diffusion process creating a much more dense frost structure. The
higher frost density allows the coil to accumulate significantly
more mass of moisture (frost) before adversely affecting coil
capacity due to airflow blockage.
Measuring Coil Capacity Decrease Due to Frosting

How significant is the rate of capacity loss due to frosting?

As mentioned previously, the loss of coil capacity under
frosting operation is due to reduced airflow, as well as increased
resistance to heat transfer. The more significant of these two
factors is the capacity loss due to blockage of airflow.2,5,10,11,12
The effects of frost presenting an increased resistance to heat
transfer are significantly less important.2,13
Aljuwahel5 monitored the performance of a single 37 ton (130
kW) evaporator located in a penthouse in a low-temperature storage freezer. Additional details on the coil are given in Table 2.
The in situ performance of the unit was determined using an
extensive configuration of air-side instrumentation arranged to
measure entering and leaving conditions (air temperature and
moisture content), as well as the average velocity of air flowing
through the coil. In addition, data was collected to determine the
volume flow rate of air being conveyed by the units five fans.
Figure 7 shows the average face velocity of air across the coil
during frosting operation over a 41 hour period. The average
velocity of air across the frost-free coil is approximately 560 ft/
min (2.85 m/s) but that average velocity decreases by nearly 50%
to 315 ft/min (1.6 m/s) at the end of its operating cycle. Figure 8
shows the average dry-bulb temperature of air entering and leaving the evaporator during frosting operation. The average entering
air temperature (i.e., space temperature) is relatively constant at
17.5F (28C) while the leaving temperature decreased from
24F (31C) to 26F (33C) as the coil accumulated frost.
The drop in leaving air temperature is a byproduct of the decreased
airflow rate through the coil, which allows longer dwell time to
give up its heat to the refrigerant. Unfortunately, that decreased
coil leaving air temperature is not sufficient enough to overcome
the drop in airflow rate. Consequently, the coils refrigeration
capacity decreases over time as frost accumulates on the coil. The
actual measured gross capacity of the coil is shown in Figure 9.
The average clean coil capacity over four separate runs is 33
tons (116 kW) and the capacity of the unit decreases to 27 tons
(95 kW) after 41 hours of operation, representing a capacity loss
of nearly 20%. Two other observations were made regarding the
measured coil capacity. First, the field-measured capacity is 8%
less than the units rated capacity. Second, the measured evapora32

ASHRAE Journal

Figure 6: Higher density frost forming on an evaporator.

Parameter

Value

Fin Pitch

3 fins/in. (0.85 cm)

Face Area

88.6 ft2 (8.23 m2)

Tube Diameter

3/4 in. (19.05 mm)

Tube Length

18 ft (5.5 m)

Number of Fans

Fan Power at 30F (34C)

Air Temperature

3.125 HP (2.33 kW)

Rated cfm

60,000 cfm (1,699 m3/min)

Number of Tube Rows

10

Saturated Evaporator
Temperature

30F (34.4C)

Coil Temperature Difference

10F (5.6C)

Rated Coil Capacity

37 tons (130 kW)

Fin and Tube Material

Aluminum

Evaporator Coil Type

CPR-fed Liquid Overfeed

Table 2: Geometry and operating conditions of the experimentally

monitored air-cooling evaporator.

tor capacity is gross because it does not include fan heat gains.
The net effect is that an evaporators capacity, while operating
under frosting conditions will decrease and the systems operating efficiency suffers as a result. To counter these effects, the
accumulated frost must be removed from the evaporator surface
on either a continuous or intermittent basis.
Alternative Approaches

Are there other approaches that can further reduce or eliminate the need for defrosting evaporators?
The short answer to this question is not really. Some alternative approaches use a liquid desiccant media such as glycol, which
is sprayed directly onto the evaporator surface to preferentially
absorb the moisture into the freezing point depressed working
fluid. As moisture from the air goes into the liquid solution, the
concentration of glycol will be reduced and reconcentration
becomes necessary to avoid freeze-ups. In this case, the equivalent to a hot gas defrost for a typical evaporator occurs remotely
from the unit as heat is added to drive off the accumulated water;
thereby, re-concentrating the glycol for reuse.
Another alternative that has been promoted to reduce latent
loads is the use of solid desiccants. The solid desiccant system
a s h r a e . o r g

February 2009

References

Run #1
Run #2
Run #3
Run #4
Run #5

3.00
2.50
2.00
1.50

Region of Experiment Uncertainty

1.00
0.50
0.00
0

250

500

750

1000

1250
1500
Time (min)

1750

2000

2250

2500

Figure 7: Average face velocity of air across the coil during frosting operation.
11

24
25
26

Average Inlet Air Temperature

Run #2 Run #4
Run #3 Run #5
Model Prediction

13
15

Average Air Temperature (F)

With the exception of sprayed desiccant units,

evaporators operating at lower temperature conditions
will result in frost formation on the coil surface. Many
factors influence the rate and nature of frost formation
including: evaporator unit fin spacing, coil location,
latent (moisture) load, and frost type or structure. The
accumulation of frost on a coil causes its capacity to
decrease due to blockage of airflow, as well as the
insulating effects of the frost layer itself. As a result,
periodic removal of the accumulated frost layer is
required to maintain system capacity and efficiency.

3.50

Air Velocity (m/s)

Conclusions

4.00

Average Air Temperature (C)

approach can reduce latent loads in temperature

controlled spaces but the added cost of the desiccant system operation must be carefully evaluated
to understand whether or not the total cost of
operation will be lowered.

February 2009

Cooling Load (Ton)

Cooling Load (kW)

1. 2006 ASHRAE HandbookRefrigeration, Chapter

27
17
42, Forced-Circulation Air Coolers.
28
2. Stoecker, W.F. 1957. How frost formation on coils
19
affects refrigeration systems. Refrigerating Engineer29
21
ing, 65(2):4445.
Average Outlet Air Temperature
30
3. Cleland, D.J. 2005. Implications of coil frosting
23
31
on system designs for low-temperature applications.
25
ASHRAE Transactions, 111 (1):336345.
32
Run #2 Run #4
4. Mago, P.J. and S.A. Sherif. 2005. Coil frosting and
27
Run #3 Run #5
33
defrosting issues at low freezer temperatures near saturaModel Prediction
34
29
tion conditions. ASHRAE Transactions, 111(1):317.
0
250
500
750
1000 1250 1500 1750 2000 2250 2500
5. Aljuwayhel, N.F., D.T. Reindl, S.A. Klein, G.F.
Time (min)
Nellis. 2008. Experimental investigation of the performance of industrial evaporator coils operating under Figure 8: Average coil inlet and outlet temperatures during frosting operation.
frosting conditions. International Journal of Refrigera40
tion, 31(1):98106.
6. Smith, G.R. 1989. Theoretical cooling coil calcuRun #1 Run #3
lations at freezer temperatures to avoid unfavorable coil
126
35
Run #2 Run #4
frost. ASHRAE Transactions, 95(2):11381148.
Model Prediction
7. Smith, G.R. 1992. Latent heat, equipment-related
108
30
load, and applied psychrometrics at freezer temperatures. ASHRAE Transactions, 98(2):649657.
90
8. Reindl, D.T. and T.B. Jekel. 2008. Infiltration rate
25
determination for low temperature freezing systems.
ASHRAE Transactions, 114(2):264272.
72
20
9. IRC. 2005. Air balance issues in refrigerated
facilities. Cold Front. Industrial Refrigeration Consor54
15
tium, University of Wisconsin-Madison, 5(4):18.
10. Barrow, H. 1985. A note on frosting of heat
0
250
500
750
1000 1250 1500 1750 2000 2250 2500
pump evaporator surfaces. Heat Recovery Systems,
Time (min)
5(3):195201.
11. Seker, D., H. Karatas, and N. Egrican. 2004. Figure 9: Coil capacity (load) as the unit operates from clean to frosted condition.
Frost formation on fin-and-tube heat exchangers. Part
Imodeling of frost formation on fin-and-tube heat exchangers.
13. Machielsen, C.H. and H.G. Kerschbaumer. 1989. Influence
International Journal of Refrigeration, 27: 367374.
of frost formation and defrosting on the performance of air coolers:
12. Yao, Y., Y. Jiang, S. Deng, and Z. Ma. 2004. A study on the standards and dimensionless coefficients for the system designer.
performance of the airside heat exchanger under frosting in an air International Journal of Refrigeration, 12(3):283290.
14. 2006 ASHRAE HandbookRefrigeration, Chapter 13, Refrigsource heat pump water heater/chiller unit. International Journal of
eration Load.
Heat and Mass Transfer.
ASHRAE Journal

33